"Empathetic leadership" became one of the most-deployed phrases in business writing during the pandemic and one of the most-emptied of meaning since. The argument that survives the cycle is narrower and more useful than the slogan: empathy as a measurable management skill — distinct from sympathy, distinct from charisma, distinct from being a "nice manager" — is one of the strongest predictors of team retention and discretionary effort that the organizational-psychology literature has produced in twenty years. The reason most companies have not benefited from this is that they trained empathy as a sentiment rather than a practice.
The empirical case starts with Gallup, whose State of the Global Workplace report tracks employee engagement across roughly 160 countries. Their multi-decade finding is unchanged: about 70% of the variance in team engagement is attributable to the direct manager. The single behavioral cluster that distinguishes managers whose teams stay versus leave is what Gallup calls "individualization" — knowing each direct report's strengths, life context, and motivators well enough to manage them as a person rather than a role. That is empathy operationalized.
The skill is teachable; the sentiment is not the point
One of the most useful research streams here belongs to Helen Riess at Massachusetts General Hospital, whose work on the E.M.P.A.T.H.Y. clinical training program — published in Academic Medicine and elsewhere — has shown measurable improvements in patient-rated empathy after relatively short structured training. The transfer to management is direct. Riess's framework is built around specific observable behaviors: eye contact, posture, vocal tone calibration, naming of the other person's stated affect, perspective-taking statements. Those are skills, and they respond to deliberate practice the way other skills do.
Daniel Goleman's older work on emotional intelligence ("What Makes a Leader," Harvard Business Review, 1998, repeatedly updated) is the popular reference, but the more rigorous current literature treats empathy as a measurable competency cluster with at least three dimensions: cognitive empathy (correctly modeling another's perspective), emotional empathy (feeling the affective state), and compassionate empathy (translating both into action). Managers can be high on one and weak on the others — and the dimension that matters most for retention is cognitive empathy, not emotional.
The data behind why this is no longer optional
Gallup's 2024 State of the Global Workplace put global employee engagement at 23% — meaning roughly three-quarters of the world's workforce is either not engaged or actively disengaged at work — and estimated the global productivity cost at $8.9 trillion, around 9% of global GDP. The same report identifies that managers themselves are increasingly worse off than their reports: manager engagement dropped faster than non-manager engagement in 2023, and burnout among first-line managers reached its highest level in the survey's history. The implication is that the burden of "empathy" cannot fall on already-burned-out middle managers without something else giving.
Christina Maslach's burnout research — most accessible in The Burnout Challenge (Harvard, 2022, with Michael Leiter) — identifies six structural conditions that produce burnout: workload, control, reward, community, fairness, and values. None of those are about whether the manager sends thoughtful Slack messages. They are about whether the system the manager operates inside is designed to be survivable. Empathy without those structural conditions becomes another expectation loaded onto an already overloaded role.
The empathy-without-structure trap
This is the most important honest qualification. Companies that exhort managers to be empathetic while raising span of control, cutting administrative support, and shrinking promotion budgets are not asking for empathy; they are asking for emotional labor as unpaid compensation for structural decay. Sylvia Ann Hewlett's Center for Talent Innovation work documented for years that women and managers of color disproportionately absorb this labor, and it disproportionately fails to be recognized in performance reviews. That is the ethical cost of treating empathy as a moral virtue rather than a designed practice.
What empathy looks like as a practice, not a poster
Three behaviors have unusually strong evidence. The first is the regular one-on-one with a fixed agenda that the report — not the manager — owns. Microsoft's internal research on this, since codified in their People Analytics work, found that managers who held weekly or biweekly 1:1s of at least 30 minutes had teams with measurably higher retention and lower attrition six and twelve months out. The second is named perspective-taking: literally saying back, "Here's what I'm hearing — am I getting that right?" before responding. The third is the practice of asking about life context before asking about deliverables, particularly in moments of visible stress.
None of these is mysterious. They are also not free: they cost manager time and they require organizational tolerance for not converting every conversation into a productivity transaction. Where companies have implemented them well, the return is large and measurable in retention; where they have implemented the language without the time, the result is the cynical pattern Gallup documents — high stated empathy scores from leaders and falling engagement scores from workers.
The connection to the broader workforce moment
The reason empathy is in the conversation now is partly the pandemic, but more durably the structural fact that workers have more outside options than they did a decade ago, more information about what management practice looks like elsewhere, and lower switching costs in many sectors. McKinsey's Great Attrition studies in 2022 found that the top reasons workers gave for leaving were not compensation but "not feeling valued by the organization" and "not feeling valued by my manager." Those are empathy variables. They are also retention variables, with measurable dollar cost. The case for empathy is not moral; it is a balance-sheet case that happens to align with the moral one.
This connects directly to NWLB's research on The Burnout Decade →, which treats burnout as a structural-design problem rather than an individual-resilience problem. The two arguments are the same argument: stop asking workers and managers to absorb design flaws with personal virtue, and start designing the system to work.
What to actually do on Monday
Audit your manager span of control: if any first-line manager has more than 8–10 direct reports, the math against meaningful 1:1s defeats anything you put on a training slide. Make the weekly 1:1 a non-cancellable calendar item, with a reports-owned agenda. Train perspective-taking as a specific verbal behavior rather than a virtue. Stop measuring "empathetic culture" through employee-engagement surveys alone and start measuring it through retention by manager. And finally: pay the managers for it. Recognition, promotion, and compensation are downstream of what gets measured; empathy will be performed as theater until it is rewarded as a job.
Empathy is not a feeling you train into managers. It is a set of behaviors you make time for — and an organizational design that lets those behaviors survive the quarter.
Updated May 21, 2026. This piece was substantively rewritten as part of NWLB's 2026 editorial refresh.
